Nvidia backs Intel with $5 billion investment, but stops short of foundry deal
Chip giant backs struggling rival, joins forces on PC and data center chips, but leaves manufacturing business in limbo.
Nvidia on Thursday said it will invest $5 billion in Intel, throwing its weight behind the struggling U.S. chipmaker but stopping short of awarding Intel a crucial manufacturing contract.
The pact, which also includes plans for Intel and Nvidia to jointly develop PC and data center chips, poses a potential risk to Taiwan’s TSMC. TSMC currently manufactures Nvidia’s flagship processors, business that could one day shift to Intel. AMD, which competes with Intel in supplying data center chips, also stands to lose from Nvidia’s backing.
Nvidia, whose chips power much of the global artificial intelligence boom, said in a statement it will pay $23.28 per share for Intel common stock, slightly below Intel’s Wednesday close of $24.90 but higher than the $20.47 price the U.S. government paid for a 10% stake in Intel last month.
Nvidia is set to become one of Intel's largest shareholders, likely owning 4% or more of the company once new shares are issued.
The investment marks a new opening for Intel after years of turnaround efforts failed to gain traction. The company, once the industry’s undisputed leader that put the “silicon” in Silicon Valley, appointed a new CEO, Lip-Bu Tan, in March. Tan has pledged to slim down operations and expand capacity only when demand justifies it.
Crucially, the deal does not involve Intel’s contract manufacturing business, known in the industry as a foundry, making chips for Nvidia. Analysts have long believed that for Intel’s foundry unit to survive, it will eventually need to win a major customer such as Nvidia, Apple, Qualcomm, or Broadcom.
Still, the deal bolsters Intel’s balance sheet, weeks after it announced a $2 billion investment from SoftBank and a $5.7 billion cash infusion from the U.S. government. Intel CFO David Zinsner recently told investors the company was in a “good cash position” and would not need much more capital until it begins heavy investment in 14A, its next-generation manufacturing process.
Under the partnership, Intel will design custom data center CPUs that Nvidia will package with its AI GPUs. A proprietary Nvidia technology will enable the Intel and Nvidia chips to communicate at higher speeds than before.
These faster connections are critical in AI, where chips must be linked together to process vast amounts of data. Nvidia’s best-selling AI servers with these high-speed links are currently limited to Nvidia-only chips, but the deal puts Intel on equal footing, allowing it to profit from each Nvidia server sold.
The combined Nvidia-Intel chips could create a formidable rival to AMD, which is developing its own AI servers, and Broadcom, which provides chip-to-chip connectivity and helps design AI processors for companies such as Google.
For consumer PCs, Nvidia will provide Intel with a custom graphics chip to pair with its CPUs, using the same high-speed links, potentially giving Intel an advantage over AMD.
Although Intel’s x86 architecture has lost share to Arm-based chips in both PCs and data centers, it still dominates overall market share.
“This historic collaboration tightly couples Nvidia's AI and accelerated computing stack with Intel’s CPUs and the vast x86 ecosystem, a fusion of two world-class platforms,” Nvidia CEO Jensen Huang said. “Together, we will expand our ecosystems and lay the foundation for the next era of computing.”
The companies did not disclose financial terms of their technical collaboration but said they will develop “multiple generations” of future products. Nvidia and Intel described the arrangement as a commercial partnership, with both supplying chips to one another, and stressed there is no licensing component.
No timeline was given for the first joint products, though both companies said their existing product roadmaps remain unchanged.
Nvidia in recent years has entered both the PC and data center CPU markets, while Intel has launched several AI chips of its own in an effort to compete with Nvidia.
